Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

Poster for Zero-Shot Learning of Intent Embeddings for Expansion by Convolutional Deep Structured Semantic Models

CDSSM for Zero-Shot Intent Modeling at ICASSP16
Abstract: 

The recent surge of intelligent personal assistants motivates spoken language understanding of dialogue systems. However, the domain constraint along with the inflexible intent schema remains a big issue. This paper focuses on the task of intent expansion, which helps remove the domain limit and make an intent schema flexible. A convolutional deep structured semantic model (CDSSM) is applied to jointly learn the representations for human intents and associated utterances. Then it can flexibly generate new intent embeddings without the need of training samples and model-retraining, which bridges the semantic relation between seen and unseen intents and further performs more robust results. Experiments show that CDSSM is capable of performing zero-shot learning effectively, e.g. generating embeddings of previously unseen intents, and therefore expand to new intents without re-training, and outperforms other semantic embeddings. The discussion and analysis of experiments provide a future direction for reducing human effort about annotating data and removing the domain constraint in spoken dialogue systems.

up
0 users have voted:

Paper Details

Authors:
Dilek Hakkani-Tur, Xiaodong He
Submitted On:
31 March 2016 - 7:30pm
Short Link:
Type:
Poster
Event:
Presenter's Name:
Yun-Nung Chen
Paper Code:
HLT-P1.1
Document Year:
2016
Cite

Document Files

ZeroShot_poster.pdf

(191 downloads)

Subscribe

[1] Dilek Hakkani-Tur, Xiaodong He, "Poster for Zero-Shot Learning of Intent Embeddings for Expansion by Convolutional Deep Structured Semantic Models", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1078. Accessed: Aug. 22, 2017.
@article{1078-16,
url = {http://sigport.org/1078},
author = {Dilek Hakkani-Tur; Xiaodong He },
publisher = {IEEE SigPort},
title = {Poster for Zero-Shot Learning of Intent Embeddings for Expansion by Convolutional Deep Structured Semantic Models},
year = {2016} }
TY - EJOUR
T1 - Poster for Zero-Shot Learning of Intent Embeddings for Expansion by Convolutional Deep Structured Semantic Models
AU - Dilek Hakkani-Tur; Xiaodong He
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1078
ER -
Dilek Hakkani-Tur, Xiaodong He. (2016). Poster for Zero-Shot Learning of Intent Embeddings for Expansion by Convolutional Deep Structured Semantic Models. IEEE SigPort. http://sigport.org/1078
Dilek Hakkani-Tur, Xiaodong He, 2016. Poster for Zero-Shot Learning of Intent Embeddings for Expansion by Convolutional Deep Structured Semantic Models. Available at: http://sigport.org/1078.
Dilek Hakkani-Tur, Xiaodong He. (2016). "Poster for Zero-Shot Learning of Intent Embeddings for Expansion by Convolutional Deep Structured Semantic Models." Web.
1. Dilek Hakkani-Tur, Xiaodong He. Poster for Zero-Shot Learning of Intent Embeddings for Expansion by Convolutional Deep Structured Semantic Models [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1078